Most ankle mobility drills are calf stretches with a new name, and the best review of calf stretching found it buys a few degrees of dorsiflexion at most. That finding sits at the top of this page rather than the bottom, because everything below it has to be read against it.
What follows is the eight ankle movements in Limber's catalogue, ordered by what they load, with the honest grade attached to each idea: what has been tested, what has not, and what to do with the difference.
The short answer
- Static calf stretching raises ankle dorsiflexion by only a few degrees, of questionable clinical importance. That is a systematic review, not one small trial.
- Loaded work in the range — driving the knee over the toe with the heel down, time in the bottom of a squat — is the better bet mechanically, and it is less well tested, not better tested.
- Knee straight and knee bent load different muscles. Do both.
- The one study that moved the number properly used an hour a day with a splint. Note the dose before you take it as encouragement.
- Measure with the knee-to-wall before and after, three weeks apart, both sides. A one-centimetre change is inside the test's own error.
What calf stretching actually does
Radford and colleagues reviewed the trials of static calf stretching in the British Journal of Sports Medicine in 2006. Pooled, stretching produced a statistically significant increase in ankle dorsiflexion — of a few degrees, which the review itself called of questionable clinical importance.
Grade: solid, and it goes against this page. The effect is real and small. Anyone promising a transformed ankle from a wall calf stretch is selling something the literature does not support.
The prevention claim is worse. Pope and colleagues randomised 1,538 army recruits to calf stretching before exercise in Medicine & Science in Sports & Exercise in 2000 and found no meaningful reduction in injury.
Grade: solid, and it goes against this page too. Stretching the calf is not injury insurance, and no page here will imply it is.
So why stretch at all? Because a few degrees is not nothing when you are near the floor of the scale, because the movements are cheap and low risk, and because the alternative — doing nothing and assuming the ankle is fine — is worse. But hold the expectation at the size the evidence supports, and put the number on paper so you can check.
The ankle mobility drills
Eight movements, which is the whole of Limber's calves and ankles catalogue. They are listed by what they load rather than by how they look, because that is the only ordering that helps you choose.
| Move | What it loads | How it runs |
|---|---|---|
| M053 Seated Ankle CARs | The joint itself, every direction | Dynamic, 40 s, per side, nothing needed |
| M011 Wall Ankle Dorsiflexion Rock | The loaded direction — knee over toe | Dynamic, 40 s, per side, a wall |
| M007 Downward Dog Calf Pedals | Both calves alternately, plus hamstrings | Dynamic, 45 s, a mat |
| M055 Heel Walks | The front of the shin — the other side | Dynamic, 40 s, nothing needed |
| M051 Wall Calf Stretch, Straight Knee | Gastrocnemius, the knee-crossing muscle | Hold, 40 s, per side, a wall |
| M052 Wall Calf Stretch, Bent Knee | Soleus and the back of the joint | Hold, 40 s, per side, a wall |
| M054 Toe-Up Wall Calf Hold | The calf line down into the arch | Hold, 40 s, per side, a wall |
| M056 Kneeling Toe Stretch | The toes and the sole of the foot | Hold, 30 s, a mat |
A few notes the table cannot carry.
M011 Wall Ankle Dorsiflexion Rock is the closest thing here to the test itself: heel planted, knee driven forward over the toes and back, rhythmically. It is cleared for warm-ups, and its own instruction is to ease off at pinching at the front of the ankle — which is the signal that you have stopped stretching tissue and started compressing a joint.
M051 and M052 are a pair, not alternatives. Straight knee puts the gastrocnemius on stretch because it starts above the knee. Bent knee slackens it and leaves the soleus and the back of the joint. M051 is the one movement here Limber will run as PNF; M052 is not.
M055 Heel Walks loads the front of the shin rather than the back of the calf. It is in the list because a joint that only ever gets stretched in one direction is being trained by half.
M056 Kneeling Toe Stretch is the only one that reaches the toes and the sole. It is a 30-second hold and it is uncomfortable in a way the others are not; take weight on your hands to ease it, and skip it if kneeling bothers your knees.
Loaded work in the range
The mechanically better idea is to spend time in the position under load rather than at the end of it under tension.
That means the knee-over-toe rock with the heel down, holding the bottom of a squat, and slow controlled work through the range rather than a long passive hold at the end of it. The reasoning is straightforward: the closed-chain movement is the one you actually use, it loads the joint as well as the muscle, and it trains the position rather than the tissue in isolation.
Grade: thin, and honestly so. This is a mechanical argument with far less trial evidence behind it than static stretching has. It is not that loaded work has been tested and won. It is that it has barely been tested at all, while stretching has been tested and produced a few degrees. Preferring it is a judgement, and this page is labelling it as one rather than dressing it up.
Afonso and colleagues found in Healthcare in 2021 that strength training and stretching produced comparable range gains, across a small and varied set of trials — which is the nearest thing to support the loaded case has, and it is not much.
The dose, and one exception
The dose research on stretching is mostly hamstrings, and it is worth borrowing carefully.
Bandy and Irion, in Physical Therapy in 1994, found 30 seconds beat 15 and matched 60 over six weeks. Their 1997 follow-up found one 30-second hold a day matched three. Behm and colleagues, in Applied Physiology, Nutrition, and Metabolism in 2016, put the useful threshold at around 60 seconds per muscle in a session. Thomas and colleagues argued in the International Journal of Sports Medicine in 2018 that total weekly time is the variable that matters, not how you cut it up.
Grade: solid for hamstrings in young healthy adults, and a reasonable default elsewhere. None of it was done on calves specifically.
Then the exception, which is instructive precisely because of its dose. Warneke and colleagues, in Frontiers in Physiology in 2022, had people stretch the calf for an hour a day using a splint, and did raise both range and muscle thickness. One laboratory, one muscle, a device, and an hour a day.
Grade: moderate for what it shows, and read the dose. It is evidence that calf muscle will change under a large enough stimulus. It is not evidence that forty seconds against a wall will do it, and quoting it as if it were is dishonest.
How to run a session
Limber's session grammar is the same here as everywhere: a warm-up of one or two dynamic movements for 60 to 90 seconds, a main block, and a downshift — one calm long hold, 45 to 60 seconds, at the end.
For ankles that comes out as something like this. Warm up with M053 Seated Ankle CARs or M055 Heel Walks. Work with M011 Wall Ankle Dorsiflexion Rock and M007 Downward Dog Calf Pedals. Hold with M051 and M052, both sides, forty seconds each. Downshift with M054 or M056.
Holds run 20 to 60 seconds. Dynamic movements count one repetition per breath cycle, six to twelve of them. Breathe in for about four seconds and out for about six. The target sensation is a mild stretch, never sharp pain.
If you use PNF, use it only on M051 — it is the one movement in this group Limber marks as capable of it — and never in a warm-up. The pattern is: settle into the hold, contract at about 40% effort for six seconds, release for two, then deepen for 15 to 20 seconds. Two or three rounds.
Ease off at a pinch in the front of the ankle rather than pushing through it. Pain that persists, wakes you at night, radiates into a limb, comes with numbness or weakness, or followed a fall or an injury is a matter for a clinician who can examine you — as is an ankle you cannot weight, an ankle that gave way, or a calf that is swollen, hot or noticeably larger than the other one. Nothing here diagnoses anything and nothing here is treatment.
Measure before and after
This is the part that makes the rest worth doing, and it is the part everyone skips.
Take the knee-to-wall test before you start. Both sides, centimetres, heel flat, foot square to the wall. Write the two numbers down with the date. Then work for three weeks and take it again in the same conditions — same hour, same footwear or none, same tape anchored the same way.
Read the second reading against the first with the error in mind. On Limber's scale a centimetre is worth five to ten points, and the published minimal detectable change for this test is about a centimetre to two. So a one-centimetre gain is not a gain. It is the instrument. Two centimetres, on both sides, three weeks apart, is a result.
If a squat is what you are after, take the deep squat test alongside it, so you can see whether the ankle change moved the movement above it. The protocol and the full scale are in the knee-to-wall test; what the reading means is in ankle dorsiflexion explained; and if the complaint is a sensation rather than a number, calf tightness separates the four things that word covers. The cluster is hips and legs.
Questions
Do ankle mobility drills actually work?
They produce a small, real change. The best systematic review of static calf stretching found an increase in ankle dorsiflexion of a few degrees, which the authors themselves described as of questionable clinical importance. Loaded work in the range is the better mechanical argument and has much less evidence behind it. Expect a modest change over weeks, measure it, and do not expect a transformation.
How long should I hold a calf stretch?
Around 30 to 40 seconds is the sensible default, and Limber's own calf holds run 40 seconds. The dose research behind that number was done on hamstrings in young healthy adults, where 30 seconds beat 15 and matched 60, and one hold a day matched three. Total time across the week matters more than the length of any single hold.
Should I stretch with the knee straight or bent?
Both, because they load different muscles. The gastrocnemius starts above the knee, so a straight-knee stretch reaches it; the soleus starts below the knee and is reached with the knee bent. Limber keeps them as a pair for exactly that reason, and a routine that only ever does one of them is training half the calf.
How often should I do them?
Most days is reasonable, and daily is not necessary. Total weekly time appears to be the variable that matters rather than the frequency, so four sessions of five minutes and seven sessions of three are much the same. Range gains also reverse after you stop, which is an argument for a small permanent habit over a short intense block.
Will these help my squat depth?
Possibly, if the ankle is what is limiting you, and a squat on its own cannot tell you that — it sees ankles, knees, hips and lower back at once. Restricting ankle dorsiflexion in healthy people changes squat mechanics immediately in controlled experiments, so the link is real in that direction. Measure the ankle first and you will know whether it is your limit before spending six weeks on it.
Can I do these before running or lifting?
The dynamic ones, yes — ankle circles, heel walks and the wall rock are all cleared as warm-up movements in Limber. The long static holds are better placed after the session or on their own, and PNF work is never a warm-up. Static stretching inside a full warm-up has not been shown to impair performance, but there is no reason to spend warm-up time on it either.